Image evaluation with Response Evaluation Criteria in Solid Tumors (RECIST) evaluates the change in a measurable lesion as determined by ruler or micrometer caliper. However, there is no definition of the conditions thought to influence the precision of measurement. We therefore examined the effects on measurement precision by changing image amplification, WW, WL, and time phase. Moreover, to determine response rate, one-dimensional evaluation with RECIST was compared with the two-dimensional evaluation of World Health Organization (WHO) for a hepatocellular carcinoma. The results of measuring the object lesion for measured value variation were as follows. Under image conditions of 1 time expansion/ WW 150/ WL 100 was (4.92+/-1.94)%. Under image conditions of 1 time expansion/ WW 350/ WL 75 was (4.42+/-1.70)%. Under image conditions of 4 times expansion/ WW 150/ WL 100 was (2.52+/-0.82)%. Under image conditions of 4 times expansion/ WW 350/ WL 75 was (2.83+/-1.10)%. When an image was enlarged to 4 times, precision doubled. There was no a difference in comparing RECIST to WHO in terms of response rate. Thus the best method was considered to be RECIST because of its convenience.